Artificial
intelligence
in
the
healthcare
field
is
full
of
promise
but
still
under-adopted,
according
to
Chirag
Shah,
partner
at
Define
Ventures.
Last
month,
the
venture
capital
firm
published
its
AI
thesis,
arguing
that
the
healthcare
industry
must
start
moving
beyond
narrow
use
cases
for
AI
and
embrace
more
workflow-integrated
platforms
in
order
to
achieve
lasting
impact.
A
convergence
of
factors
—
ongoing
money
problems,
rapid
technical
advances
and
growing
readiness
among
healthcare
customers
—
is
giving
way
to
what
Define
calls
a
“once-in-a-generation
moment”
in
which
the
best
workflow-integrated
AI
startups
can
transform
how
care
is
delivered,
paid
for
and
experienced.
Define
uses
the
“house
of
healthcare”
as
a
framework
for
understanding
healthcare
innovation.
This
includes
the
front
door,
where
patients
first
interact
with
the
system;
the
foundation,
made
up
of
data
and
infrastructure;
and
the
rooms,
representing
care
delivery.
When
it
comes
to
the
front
door,
AI
can
make
outreach
and
engagement
more
personalized
by
combining
clinical
and
personal
data.
Innovation
for
the
foundation
has
historically
centered
on
digitization
and
aggregation
—
with
AI,
healthcare
organizations
are
turning
that
data
into
insight,
Shah
explained.
As
for
the
rooms,
AI
is
already
starting
to
offload
administrative
tasks
such
as
charting,
documentation
and
messaging
so
providers
can
focus
more
on
their
patients,
he
said.
Define’s
portfolio
companies
span
all
areas
of
the
house,
Shah
stated.
One
of
these
startups
is
Luminai,
which
uses
AI
to
automate
routine
tasks
like
patient
intake,
eligibility
checks
and
documentation,
freeing
up
healthcare
staff
to
focus
on
direct
patient
care.
Another
is
Layer
Health,
which
sells
an
AI
engine
to
quickly
abstract
and
organize
clinical
data
from
charts.
There’s
a
lot
to
be
excited
about
in
terms
of
the
future
of
AI
in
the
field,
Shah
noted,
saying
that
the
technology
is
still
in
the
early
stages
of
demonstrating
its
full
potential.
As
innovation
continues,
he
believes
the
most
successful
AI
startups
will
be
the
ones
that
are
able
to
integrate
quickly
into
provider,
payer
and
pharma
workflows
without
creating
any
extra
burden.
Shah
added
that
while
it’s
easier
than
ever
to
build
a
point
solution,
it
is
much
wiser
for
startups
to
expand
into
second,
third
and
fourth
use
cases
with
customers,
evolving
their
tools
from
wedges
into
platforms.
As
he
sees
it,
companies
that
only
solve
one
narrow
pain
point
risk
being
displaced.
Portfolio
company
Cohere
Health
is
a
good
example
of
a
startup
that
expanded
the
capabilities
of
its
AI.
The
company
began
with
prior
authorization
in
musculoskeletal
care
and
then
expanded
into
oncology,
cardiology,
drugs
and
software-based
models,
Shah
explained.
“In
the
world
of
AI,
when
everybody
else
can
move
just
as
fast,
if
not
faster,
than
you
can,
one
of
the
mistakes
that
we
see
is
that
people
haven’t
done
enough
of
the
customer
discovery
work
to
understand
what’s
going
to
come
next.
After
that
wedge,
what
else
are
your
customers
going
to
need?
At
some
point,
the
competition
is
going
to
come
in,
and
the
last
thing
you
want
is
for
your
wedge
to
be
your
only
product.
We
think
it’s
really
important
to
be
building
that
—
your
product
development
cycles
have
to
get
really
accelerated
now,
especially
as
compared
to
prior
years,”
he
remarked.
From
his
perspective
as
a
digital
health
investor,
Shah
thinks
the
key
to
success
in
healthcare
AI
lies
not
in
simply
developing
a
strong
product
—
startups
need
to
expand
beyond
their
initial
use
cases
and
move
faster
than
the
competition.
Photo:
z_wei,
Getty
Images